It is commonly argued that unless under nutrition is addressed in the first 2-3 years of life, the window of opportunity for improving both growth and cognitive development of children, adolescents and adults is severely limited. However, many children who experience linear growth retardation in their first 2-3 years of life experience growth recovery later in childhood. In Peru, for example, 37% of children who were stunted at age 12 (16) mo were not stunted at age 5 y. Further, the cognitive achievement of children who recovered from stunting was no different from the cognitive achievement of children who were not stunted at either time point. The causality of this association has not been determined, and the determinants of early life growth recovery are not known. The objective of this study is to understand (1) what determines recovery from early growth failure (physically stunted at 1 year of age but not at 5 years of age), and (2) how recovery from early growth failure affects (a) schooling (b) cognitive development, and (c) academic achievement through age 12 y. We propose innovative systematic analyses of recent, very rich longitudinal data on children from four low and middle income countries on three continents. We will conduct secondary analysis of the Young Lives (YL) database. YL is the largest cross-national cohort dataset on poverty and child well-being in the developing world and is ideally placed to permit examination of the long-term impact of nutritional recovery on schooling, cognitive development and academic achievement among children growing up in the 21st century over a very significant period of their lives in a range of diverse, developing country contexts. We will analyze data from all four YL countries: Ethiopia, India, Peru and Vietnam. We will examine data for each of the four rounds of data collection (when children were 1, 5, 8 and 12 years of age). The four diverse settings and large sample sizes permit an examination of consistency across countries and an exploration of the additional burdens posed by disparities within countries in class/caste, sex, race, ethnicity, level of poverty, parental schooling and region of residence. This research is highly relevant to policies regarding children's health and education. Findings will (1) improve understanding of age patterns in growth faltering and recovery in relation to schooling and outcomes in diverse contexts, and (2) inform other long-term studies on nutritional recovery including work that tests the impact of clinic-, school- and community-based interventions on cognitive function and academic achievement.